----------------------------------------------------------------------------------------------
      name:  <unnamed>
       log:  /Users/qsg999/Desktop/PSRM_replication_material/Main_analysis.log
  log type:  text
 opened on:  18 Jun 2024, 09:02:29

. 
. use "`path'/Data.dta", clear

. cd "`path'/temp"
/Users/qsg999/Desktop/PSRM_replication_material/temp

. 
. * reversing support var for abortion
. replace support = 2 if support==1 & topic==7
(1,070 real changes made)

. replace support = 1 if support==0 & topic==7
(3,581 real changes made)

. replace support = 0 if support==2 & topic==7
(1,070 real changes made)

. 
. * three category political knowledge var
. recode pol_know (0 1=1) (2 3=2) (4 5=3), g(pk3)
(25,608 differences between pol_know and pk3)

. 
. 
. ** Figure 1
. recode support (.=2), g(re_support)
(2,878 differences between support and re_support)

. 
. 
. * Infrastructure bill
. catplot treatment re_support if topic==1, recast(bar) percent(treatment) asyvar ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Infrastructure bill") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0)) scheme(plotplain)

. 
. graph save "infrastructure", replace
file infrastructure.gph saved

. 
. 
. * min wage
. catplot treatment re_support if topic==2, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Raise the minimum wage") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "min_wage", replace
file min_wage.gph saved

. 
. 
. * estate tax
. catplot treatment re_support if topic==3, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Repeal the estate tax") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "estate_tax", replace
file estate_tax.gph saved

. 
. 
. * Capital gains tax
. catplot treatment re_support if topic==4, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Capital gains tax = income tax") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "cg_tax", replace
file cg_tax.gph saved

. 
. * Border security
. catplot treatment re_support if topic==5, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Tighten border security") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "border_security", replace
file border_security.gph saved

. 
. 
. * Transgender in military
. catplot treatment re_support if topic==6, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Allow transgender in military") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "trans_mil", replace
file trans_mil.gph saved

. 
. 
. 
. * abortion
. catplot treatment re_support if topic==7, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Keep Roe v. Wade") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "Control" 2 "Treatment")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) /
> //
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "abortion", replace
file abortion.gph saved

. 
. * Vaccine mandates
. catplot treatment re_support if topic==8, recast(bar) percent(treatment) asyvar scheme(plotp
> lainblind) ///
> bar(1, color(black) fintensity(inten80)) ///
> bar(2, color(gray) fintensity(inten80)) ///
> ytitle("") ///
> ylabel(0(25)100, labcolor(gray) ) ///
> yscale(range(0 100.1) lc(gray)) ///
> title("Vaccine mandates") ///
> var1opts(label(labsize(small) labcolor(gray)) relabel(1 "No DK-Option" 2 "DK-Option")) ///
> var2opts(label(labsize(small) labcolor(gray)) relabel(1 "Oppose" 2 "Support" 3 "DK/skip")) l
> egend(cols(2) ) ///
> graphregion(margin(0.75 0.75 0.75 0.75)) plotregion(margin(0 0 0 0))

. 
. graph save "vaccine", replace
file vaccine.gph saved

. 
. 
. grc1leg2 "infrastructure" "min_wage" "estate_tax" "cg_tax" "border_security" "trans_mil" "ab
> ortion" "vaccine",  labsize(vsmall) position(6) scheme(plotplain) r(4)  ysize(5) xsize(5) le
> gend("vaccine")
-grc1leg2- working...

. 
. graph export "figure1.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure1.pdf saved as PDF format

. 
. 
. * Table D1
. reg dk i.treatment##i.pol_know i.topic i.round, cluster(respondent_id)

Linear regression                               Number of obs     =     38,480
                                                F(20, 4809)       =     127.87
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1291
                                                Root MSE          =     .24556

                            (Std. err. adjusted for 4,810 clusters in respondent_id)
------------------------------------------------------------------------------------
                   |               Robust
                dk | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------------+----------------------------------------------------------------
       1.treatment |   .2713721   .0167465    16.20   0.000     .2385413    .3042029
                   |
          pol_know |
                1  |  -.0002533   .0006342    -0.40   0.690    -.0014967    .0009901
                2  |   .0002128   .0005201     0.41   0.682    -.0008068    .0012325
                3  |     .00035   .0005934     0.59   0.555    -.0008133    .0015133
                4  |  -.0002801   .0005889    -0.48   0.634    -.0014346    .0008744
                5  |  -.0002208    .000777    -0.28   0.776     -.001744    .0013025
                   |
treatment#pol_know |
              1 1  |  -.0691184   .0195726    -3.53   0.000    -.1074896   -.0307471
              1 2  |  -.1019593   .0181026    -5.63   0.000    -.1374487   -.0664698
              1 3  |  -.1385856   .0177776    -7.80   0.000    -.1734379   -.1037333
              1 4  |  -.1624659    .017569    -9.25   0.000    -.1969093   -.1280226
              1 5  |  -.1686643   .0183509    -9.19   0.000    -.2046404   -.1326882
                   |
             topic |
                2  |  -.0266112   .0038942    -6.83   0.000    -.0342456   -.0189769
                3  |   .1361746   .0058891    23.12   0.000     .1246293    .1477199
                4  |   .0417879   .0048445     8.63   0.000     .0322904    .0512855
                5  |   .0054054   .0045653     1.18   0.236    -.0035447    .0143556
                6  |   .0126819   .0045765     2.77   0.006     .0037098     .021654
                7  |  -.0237006    .003887    -6.10   0.000    -.0313209   -.0160804
                8  |  -.0014553   .0043276    -0.34   0.737    -.0099395    .0070289
                   |
             round |
                2  |   .0040183   .0039106     1.03   0.304    -.0036482    .0116848
                3  |  -.0041438   .0036706    -1.13   0.259    -.0113398    .0030522
                   |
             _cons |   -.017225     .00379    -4.54   0.000    -.0246552   -.0097948
------------------------------------------------------------------------------------

. outreg2 using table_D1,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +)  la
> b replace
table_D1.rtf
dir : seeout

.         
. margins, dydx(treatment) at(pol_know=(0(1)5))

Average marginal effects                                Number of obs = 38,480
Model VCE: Robust

Expression: Linear prediction, predict()
dy/dx wrt:  1.treatment
1._at: pol_know = 0
2._at: pol_know = 1
3._at: pol_know = 2
4._at: pol_know = 3
5._at: pol_know = 4
6._at: pol_know = 5

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
0.treatment  |  (base outcome)
-------------+----------------------------------------------------------------
1.treatment  |
         _at |
          1  |   .2713721   .0167465    16.20   0.000     .2385413    .3042029
          2  |   .2022537   .0101008    20.02   0.000     .1824515     .222056
          3  |   .1694128   .0068741    24.65   0.000     .1559364    .1828892
          4  |   .1327865   .0060057    22.11   0.000     .1210127    .1445604
          5  |   .1089062   .0052866    20.60   0.000     .0985421    .1192703
          6  |   .1027078   .0074828    13.73   0.000     .0880382    .1173775
------------------------------------------------------------------------------
Note: dy/dx for factor levels is the discrete change from the base level.

. 
. ** Figure 2
. marginsplot, xlabel(, nogrid labsize(medsmall) labcolor(gray)) plotopts(msym(O) msize(medlar
> ge) mlcolor("black") mfcolor("black")) recast(scatter) recastci(rspike) ci1opt(color(black) 
> lwidth(medthick)) xsize(10) scheme(plotplain) title("") xtitle("# correct answers to factual
>  questions about politics") ytitle("Treatment effect", size(medlarge) color("gray")) ylabel(
> 0(.1).32, labsize(medsmall) labcolor(gray)) xscale(r(-.2 5.2))

Variables that uniquely identify margins: pol_know

. 
. graph export "figure2.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure2.pdf saved as PDF format

. 
. su sure

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
        sure |     35,594    3.498455    .6842467          1          4

. g sure_sd=(sure-`r(mean)')/`r(sd)'
(2,886 missing values generated)

. 
. * Table D2
. reg sure_sd i.treatment##i.topic i.round, cluster(respondent_id)

Linear regression                               Number of obs     =     35,594
                                                F(17, 4808)       =     212.67
                                                Prob > F          =     0.0000
                                                R-squared         =     0.1094
                                                Root MSE          =     .94394

                         (Std. err. adjusted for 4,809 clusters in respondent_id)
---------------------------------------------------------------------------------
                |               Robust
        sure_sd | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
    1.treatment |   .1555462    .028257     5.50   0.000     .1001495    .2109428
                |
          topic |
             2  |   .4598059   .0227379    20.22   0.000     .4152293    .5043826
             3  |   -.477229   .0279339   -17.08   0.000    -.5319923   -.4224658
             4  |  -.2509917   .0270531    -9.28   0.000    -.3040282   -.1979553
             5  |   .0866928   .0272449     3.18   0.001     .0332804    .1401052
             6  |   .3680468   .0251937    14.61   0.000     .3186556    .4174381
             7  |   .6565632   .0231625    28.35   0.000     .6111542    .7019723
             8  |   .3203672   .0258876    12.38   0.000     .2696156    .3711188
                |
treatment#topic |
           1 2  |  -.0902807   .0312593    -2.89   0.004    -.1515632   -.0289981
           1 3  |   .4187054   .0401398    10.43   0.000      .340013    .4973978
           1 4  |   .0878784   .0377948     2.33   0.020     .0137833    .1619735
           1 5  |    .037785   .0375331     1.01   0.314    -.0357969     .111367
           1 6  |    .037604    .034141     1.10   0.271     -.029328    .1045359
           1 7  |  -.0490894   .0312967    -1.57   0.117    -.1104453    .0122664
           1 8  |  -.0030525   .0350668    -0.09   0.931    -.0717995    .0656945
                |
          round |
             2  |  -.0152317   .0195537    -0.78   0.436    -.0535659    .0231026
             3  |  -.0272193   .0186162    -1.46   0.144    -.0637155     .009277
                |
          _cons |  -.2356789   .0235551   -10.01   0.000    -.2818576   -.1895002
---------------------------------------------------------------------------------

. outreg2 using table_D2,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +)  la
> b replace
table_D2.rtf
dir : seeout

. 
. margins, dydx(treatment) over(topic)

Average marginal effects                                Number of obs = 35,594
Model VCE: Robust

Expression: Linear prediction, predict()
dy/dx wrt:  1.treatment
Over:       topic

------------------------------------------------------------------------------
             |            Delta-method
             |      dy/dx   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
0.treatment  |  (base outcome)
-------------+----------------------------------------------------------------
1.treatment  |
       topic |
          1  |   .1555462    .028257     5.50   0.000     .1001495    .2109428
          2  |   .0652655   .0234284     2.79   0.005     .0193352    .1111958
          3  |   .5742515   .0377237    15.22   0.000     .5002958    .6482073
          4  |   .2434245   .0329909     7.38   0.000     .1787472    .3081019
          5  |   .1933312   .0290121     6.66   0.000     .1364541    .2502083
          6  |   .1931501     .02544     7.59   0.000      .143276    .2430242
          7  |   .1064567   .0200989     5.30   0.000     .0670538    .1458597
          8  |   .1524937   .0261466     5.83   0.000     .1012343     .203753
------------------------------------------------------------------------------
Note: dy/dx for factor levels is the discrete change from the base level.

. 
. * Figure 3
. mplotoffset, offset(0.1) xlabel(1 `""Infrastructure" "bill""' 2 `""Minimum" "wage""' 3 "Esta
> te tax" 4 `""Capital gains" "tax""' 5 `""Border" "security""' 6 `""Transgender in" "military
> ""' 7 "Abortion" 8 `""Covid" "vaccines""' , nogrid labsize(medsmall) labcolor(gray)) plotopt
> s(msym(O) msize(medlarge) mlcolor("black") mfcolor("black")) recast(scatter) recastci(rspike
> ) ci1opt(color(black) lwidth(medthick)) xsize(7.5) scheme(plotplain) xtitle("") ytitle("Trea
> tment effect (standardized)", size(medlarge) color("gray")) title("") ylabel(0(.2).65, labsi
> ze(medsmall) labcolor(gray)) xscale(r(0.8 8.2))

  Variables that uniquely identify margins: topic

. 
. graph export "figure3.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure3.pdf saved as PDF format

. 
. 
. 
. 
. ** Table D3
. reg support i.treatment i.topic i.round, cluster(respondent_id)

Linear regression                               Number of obs     =     35,602
                                                F(10, 4808)       =     358.70
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0731
                                                Root MSE          =     .44851

                      (Std. err. adjusted for 4,809 clusters in respondent_id)
------------------------------------------------------------------------------
             |               Robust
     support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
 1.treatment |    .010277   .0066279     1.55   0.121    -.0027168    .0232708
             |
       topic |
          2  |   -.062998   .0061201   -10.29   0.000    -.0749962   -.0509998
          3  |  -.3052471    .008448   -36.13   0.000     -.321809   -.2886852
          4  |   -.289246   .0085063   -34.00   0.000    -.3059223   -.2725697
          5  |  -.2526074    .009792   -25.80   0.000    -.2718042   -.2334106
          6  |  -.0696178   .0065664   -10.60   0.000     -.082491   -.0567446
          7  |  -.0842634   .0063186   -13.34   0.000    -.0966508   -.0718761
          8  |  -.3421936   .0076238   -44.89   0.000    -.3571397   -.3272476
             |
       round |
          2  |  -.0110996    .008586    -1.29   0.196    -.0279322    .0057329
          3  |  -.0033085   .0079478    -0.42   0.677    -.0188897    .0122728
             |
       _cons |   .8539406   .0079865   106.92   0.000     .8382835    .8695978
------------------------------------------------------------------------------

. outreg2 using table_D3,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +)  la
> b replace
table_D3.rtf
dir : seeout

. 
. reg support i.treatment##i.topic i.round, cluster(respondent_id)

Linear regression                               Number of obs     =     35,602
                                                F(17, 4808)       =     211.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0734
                                                Root MSE          =      .4485

                         (Std. err. adjusted for 4,809 clusters in respondent_id)
---------------------------------------------------------------------------------
                |               Robust
        support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
----------------+----------------------------------------------------------------
    1.treatment |    .018587   .0104653     1.78   0.076    -.0019299    .0391038
                |
          topic |
             2  |  -.0623701   .0088494    -7.05   0.000    -.0797189   -.0450212
             3  |  -.2962354   .0109382   -27.08   0.000    -.3176792   -.2747916
             4  |  -.2727035   .0115159   -23.68   0.000    -.2952799    -.250127
             5  |  -.2479035   .0135327   -18.32   0.000    -.2744339   -.2213732
             6  |  -.0650483   .0092444    -7.04   0.000    -.0831715    -.046925
             7  |  -.0828457   .0090395    -9.16   0.000    -.1005673   -.0651241
             8  |  -.3478175   .0106671   -32.61   0.000      -.36873    -.326905
                |
treatment#topic |
           1 2  |  -.0015455   .0122013    -0.13   0.899    -.0254657    .0223747
           1 3  |  -.0217228   .0173274    -1.25   0.210    -.0556924    .0122468
           1 4  |  -.0366582   .0170932    -2.14   0.032    -.0701686   -.0031478
           1 5  |  -.0100188   .0196048    -0.51   0.609    -.0484531    .0284156
           1 6  |  -.0097448   .0131104    -0.74   0.457    -.0354471    .0159575
           1 7  |  -.0031616   .0126069    -0.25   0.802    -.0278769    .0215537
           1 8  |   .0119232   .0152378     0.78   0.434    -.0179499    .0417963
                |
          round |
             2  |  -.0110941   .0085849    -1.29   0.196    -.0279243    .0057362
             3  |  -.0032637   .0079463    -0.41   0.681    -.0188421    .0123148
                |
          _cons |   .8500171    .009027    94.16   0.000     .8323201    .8677141
---------------------------------------------------------------------------------

. outreg2 using table_D3,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +)  la
> b append
table_D3.rtf
dir : seeout

. 
. margins, at(topic=(1(1)8) treatment=(0 1) )

Predictive margins                                      Number of obs = 35,602
Model VCE: Robust

Expression: Linear prediction, predict()
1._at:  treatment = 0
        topic     = 1
2._at:  treatment = 0
        topic     = 2
3._at:  treatment = 0
        topic     = 3
4._at:  treatment = 0
        topic     = 4
5._at:  treatment = 0
        topic     = 5
6._at:  treatment = 0
        topic     = 6
7._at:  treatment = 0
        topic     = 7
8._at:  treatment = 0
        topic     = 8
9._at:  treatment = 1
        topic     = 1
10._at: treatment = 1
        topic     = 2
11._at: treatment = 1
        topic     = 3
12._at: treatment = 1
        topic     = 4
13._at: treatment = 1
        topic     = 5
14._at: treatment = 1
        topic     = 6
15._at: treatment = 1
        topic     = 7
16._at: treatment = 1
        topic     = 8

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8453375    .007374   114.64   0.000      .830881    .8597939
          2  |   .7829674   .0084081    93.12   0.000     .7664838     .799451
          3  |    .549102   .0101532    54.08   0.000      .529197     .569007
          4  |    .572634   .0100946    56.73   0.000      .552844     .592424
          5  |   .5974339   .0100081    59.69   0.000     .5778134    .6170545
          6  |   .7802892   .0084522    92.32   0.000      .763719    .7968593
          7  |   .7624918   .0086829    87.82   0.000     .7454694    .7795142
          8  |     .49752   .0101989    48.78   0.000     .4775254    .5175145
          9  |   .8639244   .0074268   116.32   0.000     .8493644    .8784844
         10  |   .8000089   .0084144    95.08   0.000     .7835128     .816505
         11  |   .5459662   .0129497    42.16   0.000     .5205788    .5713536
         12  |   .5545627   .0113048    49.06   0.000     .5324002    .5767252
         13  |   .6060021   .0106459    56.92   0.000     .5851313    .6268729
         14  |   .7891313   .0089631    88.04   0.000     .7715595    .8067032
         15  |   .7779171   .0087726    88.68   0.000     .7607187    .7951155
         16  |   .5280301   .0107922    48.93   0.000     .5068724    .5491879
------------------------------------------------------------------------------

. 
. * Figure 4
. mplotoffset, offset(0.1)  xlabel(1 `""Infrastructure" "bill""' 2 `""Minimum" "wage""' 3 "Est
> ate tax" 4 `""Capital gains" "tax""' 5 `""Border" "security""' 6 `""Transgender in" "militar
> y""' 7 "Abortion" 8 `""Covid" "vaccines""' , nogrid labsize(medsmall) labcolor(gray))  recas
> t(scatter) recastci(rspike) xsize(7.5) scheme(plotplain) ylab(0.5(0.1).9, labsize(medsmall) 
> labcolor(gray)) yscale(r(0.48 0.9)) yline(0.5) xtitle("") ytitle("Policy support", size(medl
> arge) color("gray")) title("") ///
> plot1opts(msymbol(O) mcolor(black))  /// marker for first line
>  ci1opts(color(black))  ///
> plot2opts(msymbol(O) mcolor(gray)) /// marker for second line
> ci2opts(color(gray))   ///
>  legend(pos(6) col(2) title("") order(3 4) label(3 "No DK-option") label(4 "DK-option") )

  Variables that uniquely identify margins: topic treatment

. graph export "figure4.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure4.pdf saved as PDF format

. 
. * Table D4
. reg support i.treatment##i.pk3 i.round if topic==1, cluster(respondent_id)

Linear regression                               Number of obs     =      4,537
                                                F(7, 4536)        =       2.00
                                                Prob > F          =     0.0513
                                                R-squared         =     0.0031
                                                Root MSE          =     .35278

                       (Std. err. adjusted for 4,537 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0557017   .0245022     2.27   0.023     .0076655     .103738
              |
          pk3 |
           2  |   .0035116   .0209117     0.17   0.867    -.0374855    .0445087
           3  |   .0230643   .0212843     1.08   0.279    -.0186633    .0647919
              |
treatment#pk3 |
         1 2  |  -.0300326   .0290276    -1.03   0.301    -.0869409    .0268757
         1 3  |    -.06259   .0299883    -2.09   0.037    -.1213817   -.0037982
              |
        round |
           2  |  -.0090793   .0136256    -0.67   0.505    -.0357921    .0176335
           3  |   .0212098   .0125641     1.69   0.091    -.0034219    .0458416
              |
        _cons |   .8301688   .0194183    42.75   0.000     .7920994    .8682381
-------------------------------------------------------------------------------

. outreg2 using table_D4,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +)  la
> b replace
table_D4.rtf
dir : seeout

. 
.                 foreach n of numlist 2/8 {
  2. reg support i.treatment##i.pk3 i.round if topic==`n', cluster(respondent_id)
  3. outreg2 using table_D4,  word  bd(3) sd(3) rd(3) alpha(0.01, 0.05, 0.1) symbol(**, *, +) 
>  lab append
  4.                 }               

Linear regression                               Number of obs     =      4,665
                                                F(7, 4664)        =       6.22
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0073
                                                Root MSE          =      .4053

                       (Std. err. adjusted for 4,665 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0301426   .0238819     1.26   0.207    -.0166772    .0769625
              |
          pk3 |
           2  |  -.0702049   .0214733    -3.27   0.001    -.1123028   -.0281071
           3  |  -.0816552   .0224742    -3.63   0.000    -.1257152   -.0375952
              |
treatment#pk3 |
         1 2  |  -.0181278   .0298517    -0.61   0.544    -.0766513    .0403957
         1 3  |  -.0130071   .0314794    -0.41   0.679    -.0747217    .0487074
              |
        round |
           2  |   -.011632    .015187    -0.77   0.444    -.0414057    .0181417
           3  |   .0118133   .0142919     0.83   0.409    -.0162055    .0398322
              |
        _cons |   .8438013   .0195256    43.22   0.000     .8055218    .8820808
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      3,882
                                                F(7, 3881)        =      12.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0217
                                                Root MSE          =     .49277

                       (Std. err. adjusted for 3,882 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0995578   .0425108    -2.34   0.019    -.1829033   -.0162122
              |
          pk3 |
           2  |   .0128651   .0283459     0.45   0.650    -.0427092    .0684394
           3  |   .1477699   .0289104     5.11   0.000     .0910889    .2044509
              |
treatment#pk3 |
         1 2  |   .1129502   .0491188     2.30   0.022      .016649    .2092514
         1 3  |   .0918868   .0495129     1.86   0.064    -.0051869    .1889605
              |
        round |
           2  |  -.0130888    .020001    -0.65   0.513    -.0523024    .0261247
           3  |  -.0204242   .0192182    -1.06   0.288     -.058103    .0172545
              |
        _cons |   .5007587   .0268231    18.67   0.000       .44817    .5533474
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      4,336
                                                F(7, 4335)        =       2.42
                                                Prob > F          =     0.0180
                                                R-squared         =     0.0039
                                                Root MSE          =     .49531

                       (Std. err. adjusted for 4,336 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0097687   .0373939    -0.26   0.794    -.0830798    .0635425
              |
          pk3 |
           2  |   .0384201   .0282025     1.36   0.173    -.0168711    .0937114
           3  |   .0432244   .0290852     1.49   0.137    -.0137974    .1002463
              |
treatment#pk3 |
         1 2  |  -.0419693   .0436442    -0.96   0.336    -.1275342    .0435956
         1 3  |   .0266579    .044702     0.60   0.551    -.0609809    .1142968
              |
        round |
           2  |  -.0300888   .0190837    -1.58   0.115    -.0675025     .007325
           3  |   -.026325   .0181987    -1.45   0.148    -.0620038    .0093538
              |
        _cons |   .5586992   .0265066    21.08   0.000     .5067327    .6106657
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      4,511
                                                F(7, 4510)        =       0.37
                                                Prob > F          =     0.9221
                                                R-squared         =     0.0006
                                                Root MSE          =      .4899

                       (Std. err. adjusted for 4,511 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0091932   .0351825     0.26   0.794    -.0597819    .0781682
              |
          pk3 |
           2  |   .0310383    .027904     1.11   0.266    -.0236671    .0857438
           3  |   .0155743   .0288631     0.54   0.590    -.0410116    .0721601
              |
treatment#pk3 |
         1 2  |  -.0075343   .0411885    -0.18   0.855     -.088284    .0732154
         1 3  |   .0066708   .0427924     0.16   0.876    -.0772232    .0905648
              |
        round |
           2  |  -.0084998   .0184863    -0.46   0.646    -.0447421    .0277425
           3  |  -.0034232   .0176958    -0.19   0.847    -.0381155    .0312692
              |
        _cons |   .5816673    .026176    22.22   0.000     .5303494    .6329851
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      4,476
                                                F(7, 4475)        =       3.18
                                                Prob > F          =     0.0023
                                                R-squared         =     0.0052
                                                Root MSE          =     .41054

                       (Std. err. adjusted for 4,476 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .050025   .0300905     1.66   0.096    -.0089673    .1090172
              |
          pk3 |
           2  |   .0318589   .0245557     1.30   0.195    -.0162825    .0800003
           3  |   .0741565   .0247818     2.99   0.003     .0255719    .1227411
              |
treatment#pk3 |
         1 2  |  -.0422045   .0352884    -1.20   0.232    -.1113873    .0269783
         1 3  |   -.059176     .03592    -1.65   0.100     -.129597     .011245
              |
        round |
           2  |   .0344229   .0158376     2.17   0.030     .0033733    .0654724
           3  |   .0492261   .0149871     3.28   0.001     .0198439    .0786082
              |
        _cons |   .7094711   .0232139    30.56   0.000     .6639604    .7549818
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      4,651
                                                F(7, 4650)        =       1.35
                                                Prob > F          =     0.2234
                                                R-squared         =     0.0019
                                                Root MSE          =     .42083

                       (Std. err. adjusted for 4,651 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0719811   .0286454     2.51   0.012     .0158224    .1281397
              |
          pk3 |
           2  |   .0107241   .0246535     0.43   0.664    -.0376085    .0590567
           3  |   .0376994   .0251005     1.50   0.133    -.0115095    .0869084
              |
treatment#pk3 |
         1 2  |  -.0571891   .0340906    -1.68   0.093     -.124023    .0096447
         1 3  |  -.0826471   .0350841    -2.36   0.019    -.1514287   -.0138656
              |
        round |
           2  |   .0035804   .0155497     0.23   0.818    -.0269045    .0340652
           3  |  -.0046158   .0149753    -0.31   0.758    -.0339746    .0247429
              |
        _cons |   .7443932   .0229665    32.41   0.000     .6993681    .7894183
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

Linear regression                               Number of obs     =      4,544
                                                F(7, 4543)        =       6.34
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0096
                                                Root MSE          =      .4979

                       (Std. err. adjusted for 4,544 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .043022   .0352313     1.22   0.222    -.0260484    .1120924
              |
          pk3 |
           2  |   .0591469   .0280748     2.11   0.035     .0041067    .1141871
           3  |   .1292174   .0289272     4.47   0.000     .0725061    .1859288
              |
treatment#pk3 |
         1 2  |   .0004325   .0415259     0.01   0.992    -.0809783    .0818434
         1 3  |  -.0363439   .0428399    -0.85   0.396     -.120331    .0476432
              |
        round |
           2  |  -.0466701   .0187185    -2.49   0.013    -.0833676   -.0099727
           3  |  -.0427228   .0178964    -2.39   0.017    -.0778085   -.0076371
              |
        _cons |   .4539623   .0262133    17.32   0.000     .4025714    .5053532
-------------------------------------------------------------------------------
table_D4.rtf
dir : seeout

.                 
. 
.                 * Figure 5
. foreach n of numlist 1/8 {
  2. reg support i.treatment##i.pk3 i.round if topic==`n', cluster(respondent_id)
  3. margins, at(pk3=(1 2 3) treatment=(0 1)) saving("h2b_topic`n'", replace)
  4. 
. preserve
  5. use "h2b_topic`n'", clear
  6. g topic=`n'
  7. save "h2b_topic`n'", replace
  8. restore
  9. 
. }

Linear regression                               Number of obs     =      4,537
                                                F(7, 4536)        =       2.00
                                                Prob > F          =     0.0513
                                                R-squared         =     0.0031
                                                Root MSE          =     .35278

                       (Std. err. adjusted for 4,537 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0557017   .0245022     2.27   0.023     .0076655     .103738
              |
          pk3 |
           2  |   .0035116   .0209117     0.17   0.867    -.0374855    .0445087
           3  |   .0230643   .0212843     1.08   0.279    -.0186633    .0647919
              |
treatment#pk3 |
         1 2  |  -.0300326   .0290276    -1.03   0.301    -.0869409    .0268757
         1 3  |    -.06259   .0299883    -2.09   0.037    -.1213817   -.0037982
              |
        round |
           2  |  -.0090793   .0136256    -0.67   0.505    -.0357921    .0176335
           3  |   .0212098   .0125641     1.69   0.091    -.0034219    .0458416
              |
        _cons |   .8301688   .0194183    42.75   0.000     .7920994    .8682381
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,537
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8353325   .0176977    47.20   0.000     .8006364    .8700285
          2  |   .8388441   .0111749    75.07   0.000     .8169358    .8607523
          3  |   .8583968   .0117947    72.78   0.000     .8352734    .8815201
          4  |   .8910342   .0169451    52.58   0.000     .8578135    .9242549
          5  |   .8645132   .0108332    79.80   0.000     .8432748    .8857515
          6  |   .8515085   .0126314    67.41   0.000     .8267449    .8762722
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic1.dta saved

Linear regression                               Number of obs     =      4,665
                                                F(7, 4664)        =       6.22
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0073
                                                Root MSE          =      .4053

                       (Std. err. adjusted for 4,665 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0301426   .0238819     1.26   0.207    -.0166772    .0769625
              |
          pk3 |
           2  |  -.0702049   .0214733    -3.27   0.001    -.1123028   -.0281071
           3  |  -.0816552   .0224742    -3.63   0.000    -.1257152   -.0375952
              |
treatment#pk3 |
         1 2  |  -.0181278   .0298517    -0.61   0.544    -.0766513    .0403957
         1 3  |  -.0130071   .0314794    -0.41   0.679    -.0747217    .0487074
              |
        round |
           2  |   -.011632    .015187    -0.77   0.444    -.0414057    .0181417
           3  |   .0118133   .0142919     0.83   0.409    -.0162055    .0398322
              |
        _cons |   .8438013   .0195256    43.22   0.000     .8055218    .8820808
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,665
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8445646   .0173136    48.78   0.000     .8106216    .8785075
          2  |   .7743596    .012716    60.90   0.000     .7494303    .7992889
          3  |   .7629094   .0143188    53.28   0.000     .7348377    .7909811
          4  |   .8747072   .0164561    53.15   0.000     .8424455    .9069689
          5  |   .7863744   .0126016    62.40   0.000     .7616693    .8110796
          6  |   .7800449   .0147003    53.06   0.000     .7512253    .8088645
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic2.dta saved

Linear regression                               Number of obs     =      3,882
                                                F(7, 3881)        =      12.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0217
                                                Root MSE          =     .49277

                       (Std. err. adjusted for 3,882 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0995578   .0425108    -2.34   0.019    -.1829033   -.0162122
              |
          pk3 |
           2  |   .0128651   .0283459     0.45   0.650    -.0427092    .0684394
           3  |   .1477699   .0289104     5.11   0.000     .0910889    .2044509
              |
treatment#pk3 |
         1 2  |   .1129502   .0491188     2.30   0.022      .016649    .2092514
         1 3  |   .0918868   .0495129     1.86   0.064    -.0051869    .1889605
              |
        round |
           2  |  -.0130888    .020001    -0.65   0.513    -.0523024    .0261247
           3  |  -.0204242   .0192182    -1.06   0.288     -.058103    .0172545
              |
        _cons |   .5007587   .0268231    18.67   0.000       .44817    .5533474
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 3,882
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4890234    .023924    20.44   0.000     .4421187    .5359282
          2  |   .5018886   .0152154    32.99   0.000     .4720577    .5317194
          3  |   .6367933   .0162175    39.27   0.000     .6049977     .668589
          4  |   .3894657   .0351542    11.08   0.000     .3205433    .4583881
          5  |    .515281    .019271    26.74   0.000     .4774987    .5530634
          6  |   .6291224   .0195329    32.21   0.000     .5908266    .6674181
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic3.dta saved

Linear regression                               Number of obs     =      4,336
                                                F(7, 4335)        =       2.42
                                                Prob > F          =     0.0180
                                                R-squared         =     0.0039
                                                Root MSE          =     .49531

                       (Std. err. adjusted for 4,336 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0097687   .0373939    -0.26   0.794    -.0830798    .0635425
              |
          pk3 |
           2  |   .0384201   .0282025     1.36   0.173    -.0168711    .0937114
           3  |   .0432244   .0290852     1.49   0.137    -.0137974    .1002463
              |
treatment#pk3 |
         1 2  |  -.0419693   .0436442    -0.96   0.336    -.1275342    .0435956
         1 3  |   .0266579    .044702     0.60   0.551    -.0609809    .1142968
              |
        round |
           2  |  -.0300888   .0190837    -1.58   0.115    -.0675025     .007325
           3  |   -.026325   .0181987    -1.45   0.148    -.0620038    .0093538
              |
        _cons |   .5586992   .0265066    21.08   0.000     .5067327    .6106657
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,336
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5393604   .0238596    22.61   0.000     .4925834    .5861374
          2  |   .5777805   .0150441    38.41   0.000     .5482865    .6072746
          3  |   .5825848   .0166173    35.06   0.000     .5500063    .6151633
          4  |   .5295917   .0288006    18.39   0.000     .4731277    .5860557
          5  |   .5260426   .0167219    31.46   0.000      .493259    .5588261
          6  |    .599474   .0180027    33.30   0.000     .5641795    .6347686
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic4.dta saved

Linear regression                               Number of obs     =      4,511
                                                F(7, 4510)        =       0.37
                                                Prob > F          =     0.9221
                                                R-squared         =     0.0006
                                                Root MSE          =      .4899

                       (Std. err. adjusted for 4,511 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0091932   .0351825     0.26   0.794    -.0597819    .0781682
              |
          pk3 |
           2  |   .0310383    .027904     1.11   0.266    -.0236671    .0857438
           3  |   .0155743   .0288631     0.54   0.590    -.0410116    .0721601
              |
treatment#pk3 |
         1 2  |  -.0075343   .0411885    -0.18   0.855     -.088284    .0732154
         1 3  |   .0066708   .0427924     0.16   0.876    -.0772232    .0905648
              |
        round |
           2  |  -.0084998   .0184863    -0.46   0.646    -.0447421    .0277425
           3  |  -.0034232   .0176958    -0.19   0.847    -.0381155    .0312692
              |
        _cons |   .5816673    .026176    22.22   0.000     .5303494    .6329851
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,511
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5777193   .0236249    24.45   0.000      .531403    .6240357
          2  |   .6087577   .0148634    40.96   0.000     .5796181    .6378972
          3  |   .5932936   .0165586    35.83   0.000     .5608306    .6257567
          4  |   .5869125   .0260777    22.51   0.000     .5357873    .6380376
          5  |   .6104166   .0154049    39.62   0.000     .5802155    .6406177
          6  |   .6091576   .0178752    34.08   0.000     .5741134    .6442018
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic5.dta saved

Linear regression                               Number of obs     =      4,476
                                                F(7, 4475)        =       3.18
                                                Prob > F          =     0.0023
                                                R-squared         =     0.0052
                                                Root MSE          =     .41054

                       (Std. err. adjusted for 4,476 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .050025   .0300905     1.66   0.096    -.0089673    .1090172
              |
          pk3 |
           2  |   .0318589   .0245557     1.30   0.195    -.0162825    .0800003
           3  |   .0741565   .0247818     2.99   0.003     .0255719    .1227411
              |
treatment#pk3 |
         1 2  |  -.0422045   .0352884    -1.20   0.232    -.1113873    .0269783
         1 3  |   -.059176     .03592    -1.65   0.100     -.129597     .011245
              |
        round |
           2  |   .0344229   .0158376     2.17   0.030     .0033733    .0654724
           3  |   .0492261   .0149871     3.28   0.001     .0198439    .0786082
              |
        _cons |   .7094711   .0232139    30.56   0.000     .6639604    .7549818
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,476
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7386351   .0209952    35.18   0.000     .6974742    .7797959
          2  |   .7704939   .0127609    60.38   0.000     .7454762    .7955117
          3  |   .8127916   .0131665    61.73   0.000     .7869787    .8386044
          4  |     .78866   .0215784    36.55   0.000     .7463558    .8309643
          5  |   .7783144   .0133269    58.40   0.000     .7521871    .8044418
          6  |   .8036405   .0145637    55.18   0.000     .7750885    .8321925
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic6.dta saved

Linear regression                               Number of obs     =      4,651
                                                F(7, 4650)        =       1.35
                                                Prob > F          =     0.2234
                                                R-squared         =     0.0019
                                                Root MSE          =     .42083

                       (Std. err. adjusted for 4,651 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0719811   .0286454     2.51   0.012     .0158224    .1281397
              |
          pk3 |
           2  |   .0107241   .0246535     0.43   0.664    -.0376085    .0590567
           3  |   .0376994   .0251005     1.50   0.133    -.0115095    .0869084
              |
treatment#pk3 |
         1 2  |  -.0571891   .0340906    -1.68   0.093     -.124023    .0096447
         1 3  |  -.0826471   .0350841    -2.36   0.019    -.1514287   -.0138656
              |
        round |
           2  |   .0035804   .0155497     0.23   0.818    -.0269045    .0340652
           3  |  -.0046158   .0149753    -0.31   0.758    -.0339746    .0247429
              |
        _cons |   .7443932   .0229665    32.41   0.000     .6993681    .7894183
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,651
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7437796   .0208863    35.61   0.000     .7028325    .7847267
          2  |   .7545037   .0130991    57.60   0.000     .7288232    .7801843
          3  |   .7814791   .0139203    56.14   0.000     .7541886    .8087695
          4  |   .8157607    .019614    41.59   0.000      .777308    .8542134
          5  |   .7692956   .0130719    58.85   0.000     .7436686    .7949227
          6  |    .770813   .0147233    52.35   0.000     .7419483    .7996777
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic7.dta saved

Linear regression                               Number of obs     =      4,544
                                                F(7, 4543)        =       6.34
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0096
                                                Root MSE          =      .4979

                       (Std. err. adjusted for 4,544 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .043022   .0352313     1.22   0.222    -.0260484    .1120924
              |
          pk3 |
           2  |   .0591469   .0280748     2.11   0.035     .0041067    .1141871
           3  |   .1292174   .0289272     4.47   0.000     .0725061    .1859288
              |
treatment#pk3 |
         1 2  |   .0004325   .0415259     0.01   0.992    -.0809783    .0818434
         1 3  |  -.0363439   .0428399    -0.85   0.396     -.120331    .0476432
              |
        round |
           2  |  -.0466701   .0187185    -2.49   0.013    -.0833676   -.0099727
           3  |  -.0427228   .0178964    -2.39   0.017    -.0778085   -.0076371
              |
        _cons |   .4539623   .0262133    17.32   0.000     .4025714    .5053532
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,544
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4233509   .0235971    17.94   0.000      .377089    .4696128
          2  |   .4824977   .0152245    31.69   0.000     .4526503    .5123452
          3  |   .5525683   .0167066    33.07   0.000     .5198152    .5853214
          4  |   .4663729   .0261734    17.82   0.000     .4150603    .5176854
          5  |   .5259523   .0158352    33.21   0.000     .4949075    .5569971
          6  |   .5592464   .0177537    31.50   0.000     .5244404    .5940523
------------------------------------------------------------------------------
(Created by command margins; also see char list)
file h2b_topic8.dta saved

. 
. preserve

. use "h2b_topic1", clear
(Created by command margins; also see char list)

. 
. foreach n of numlist 2/8 {
  2. append using "h2b_topic`n'"
  3. }
(label _term already defined)
(label _atopt already defined)
(variable _pvalue was byte, now float to accommodate using data's values)
(label _term already defined)
(label _atopt already defined)
(label _term already defined)
(label _atopt already defined)
(label _term already defined)
(label _atopt already defined)
(label _term already defined)
(label _atopt already defined)
(label _term already defined)
(label _atopt already defined)
(label _term already defined)
(label _atopt already defined)

. 
. label define topiclab_alt2 1 "Infrastructure" 2 "Min wage" 3 "Estate tax" 4 "CG tax" 5 "Bord
> er sec" 6 "Transgender" 7 "Abortion" 8 "Covid"

. label values topic topiclab_alt2

. 
. * add a little "jitter"
. replace _at2=.90 if _at2==1 & _at1==0
variable _at2 was byte now float
(8 real changes made)

. replace _at2=1.90 if _at2==2 & _at1==0
(8 real changes made)

. replace _at2=2.90 if _at2==3 & _at1==0
(8 real changes made)

. 
. 
. replace _at2=1.1 if _at2==1 & _at1==1
(8 real changes made)

. replace _at2=2.1 if _at2==2 & _at1==1
(8 real changes made)

. replace _at2=3.1 if _at2==3 & _at1==1
(8 real changes made)

. 
. 
. twoway ///
> (scatter _margin _at2 if _at1==0, msymbol(O) mcolor(black) lcolor(black)) ///
> (scatter _margin _at2 if _at1==1, msymbol(O) mcolor(gray) lcolor(gray) lpattern(solid)) ///
> (rspike _ci_lb _ci_ub _at2 if _at1==0,  lcolor(black)  lwidth(medthick)) ///
> (rspike _ci_lb _ci_ub _at2 if _at1==1, lcolor(gray)  lwidth(medthick)), ///
> by(topic, r(1) note("") title("", size(large) position(11))) subtitle(,  size(medsmall)) ///
> scheme(plotplain) yline(.5) legend(pos(6) col(2) title("")  order(1 2) label(1 "No DK-option
> ") label(2 "DK-option")) xlabel(1 "Low" 2 "Med" 3 "High", labsize(medsmall)  labcolor(gray) 
> tlength(0.5)) xtick(1 2 3) ylabel(0.3(.1)0.9, labsize(medsmall) labcolor(gray)) xtitle("Poli
> tical knowledge", size(medsmall) color(gray)) ytitle("Policy support", size(medsmall) color(
> gray)) xsize(8)

. graph export "figure5.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure5.pdf saved as PDF format

. 
. restore

. 
. 
. * Figure 6
. preserve

. foreach n of numlist 1/8 {
  2. reg support i.treatment##i.pk3 i.round if topic==`n', cluster(respondent_id)
  3. margins, at(pk3=(1 2 3) treatment=(0 1)) post
  4. xlincom 1._at-3._at, post
  5. 
. esttab, ci
  6. 
. mat ci`n'= r(coefs)
  7. 
. reg support i.treatment##i.pk3 i.round if topic==`n', cluster(respondent_id)
  8. margins, at(pk3=(1 2 3) treatment=(0 1)) post
  9. xlincom 4._at-6._at, post
 10. 
. esttab, ci
 11. 
. mat cit`n'= r(coefs)
 12. 
. 
. reg support i.treatment##i.pk3 i.round if topic==`n', cluster(respondent_id)
 13. margins, at(pk3=(1 2 3) treatment=(0 1)) post
 14. xlincom ((4._at-6._at)-(1._at-3._at)), post
 15. 
. esttab, se
 16. 
. mat did`n'= r(coefs)
 17. 
. }

Linear regression                               Number of obs     =      4,537
                                                F(7, 4536)        =       2.00
                                                Prob > F          =     0.0513
                                                R-squared         =     0.0031
                                                Root MSE          =     .35278

                       (Std. err. adjusted for 4,537 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0557017   .0245022     2.27   0.023     .0076655     .103738
              |
          pk3 |
           2  |   .0035116   .0209117     0.17   0.867    -.0374855    .0445087
           3  |   .0230643   .0212843     1.08   0.279    -.0186633    .0647919
              |
treatment#pk3 |
         1 2  |  -.0300326   .0290276    -1.03   0.301    -.0869409    .0268757
         1 3  |    -.06259   .0299883    -2.09   0.037    -.1213817   -.0037982
              |
        round |
           2  |  -.0090793   .0136256    -0.67   0.505    -.0357921    .0176335
           3  |   .0212098   .0125641     1.69   0.091    -.0034219    .0458416
              |
        _cons |   .8301688   .0194183    42.75   0.000     .7920994    .8682381
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,537
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8353325   .0176977    47.20   0.000     .8006364    .8700285
          2  |   .8388441   .0111749    75.07   0.000     .8169358    .8607523
          3  |   .8583968   .0117947    72.78   0.000     .8352734    .8815201
          4  |   .8910342   .0169451    52.58   0.000     .8578135    .9242549
          5  |   .8645132   .0108332    79.80   0.000     .8432748    .8857515
          6  |   .8515085   .0126314    67.41   0.000     .8267449    .8762722
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0230643   .0212843    -1.08   0.279    -.0647919    .0186633
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0231   
                   [-0.0648,0.0187]   
--------------------------------------
N                              4537   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,537
                                                F(7, 4536)        =       2.00
                                                Prob > F          =     0.0513
                                                R-squared         =     0.0031
                                                Root MSE          =     .35278

                       (Std. err. adjusted for 4,537 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0557017   .0245022     2.27   0.023     .0076655     .103738
              |
          pk3 |
           2  |   .0035116   .0209117     0.17   0.867    -.0374855    .0445087
           3  |   .0230643   .0212843     1.08   0.279    -.0186633    .0647919
              |
treatment#pk3 |
         1 2  |  -.0300326   .0290276    -1.03   0.301    -.0869409    .0268757
         1 3  |    -.06259   .0299883    -2.09   0.037    -.1213817   -.0037982
              |
        round |
           2  |  -.0090793   .0136256    -0.67   0.505    -.0357921    .0176335
           3  |   .0212098   .0125641     1.69   0.091    -.0034219    .0458416
              |
        _cons |   .8301688   .0194183    42.75   0.000     .7920994    .8682381
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,537
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8353325   .0176977    47.20   0.000     .8006364    .8700285
          2  |   .8388441   .0111749    75.07   0.000     .8169358    .8607523
          3  |   .8583968   .0117947    72.78   0.000     .8352734    .8815201
          4  |   .8910342   .0169451    52.58   0.000     .8578135    .9242549
          5  |   .8645132   .0108332    79.80   0.000     .8432748    .8857515
          6  |   .8515085   .0126314    67.41   0.000     .8267449    .8762722
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0395257   .0211448     1.87   0.062    -.0019285    .0809798
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         0.0395   
                  [-0.00193,0.0810]   
--------------------------------------
N                              4537   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,537
                                                F(7, 4536)        =       2.00
                                                Prob > F          =     0.0513
                                                R-squared         =     0.0031
                                                Root MSE          =     .35278

                       (Std. err. adjusted for 4,537 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0557017   .0245022     2.27   0.023     .0076655     .103738
              |
          pk3 |
           2  |   .0035116   .0209117     0.17   0.867    -.0374855    .0445087
           3  |   .0230643   .0212843     1.08   0.279    -.0186633    .0647919
              |
treatment#pk3 |
         1 2  |  -.0300326   .0290276    -1.03   0.301    -.0869409    .0268757
         1 3  |    -.06259   .0299883    -2.09   0.037    -.1213817   -.0037982
              |
        round |
           2  |  -.0090793   .0136256    -0.67   0.505    -.0357921    .0176335
           3  |   .0212098   .0125641     1.69   0.091    -.0034219    .0458416
              |
        _cons |   .8301688   .0194183    42.75   0.000     .7920994    .8682381
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,537
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8353325   .0176977    47.20   0.000     .8006364    .8700285
          2  |   .8388441   .0111749    75.07   0.000     .8169358    .8607523
          3  |   .8583968   .0117947    72.78   0.000     .8352734    .8815201
          4  |   .8910342   .0169451    52.58   0.000     .8578135    .9242549
          5  |   .8645132   .0108332    79.80   0.000     .8432748    .8857515
          6  |   .8515085   .0126314    67.41   0.000     .8267449    .8762722
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |     .06259   .0299883     2.09   0.037     .0037982    .1213817
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1               0.0626*  
                 (0.0300)   
----------------------------
N                    4537   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,665
                                                F(7, 4664)        =       6.22
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0073
                                                Root MSE          =      .4053

                       (Std. err. adjusted for 4,665 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0301426   .0238819     1.26   0.207    -.0166772    .0769625
              |
          pk3 |
           2  |  -.0702049   .0214733    -3.27   0.001    -.1123028   -.0281071
           3  |  -.0816552   .0224742    -3.63   0.000    -.1257152   -.0375952
              |
treatment#pk3 |
         1 2  |  -.0181278   .0298517    -0.61   0.544    -.0766513    .0403957
         1 3  |  -.0130071   .0314794    -0.41   0.679    -.0747217    .0487074
              |
        round |
           2  |   -.011632    .015187    -0.77   0.444    -.0414057    .0181417
           3  |   .0118133   .0142919     0.83   0.409    -.0162055    .0398322
              |
        _cons |   .8438013   .0195256    43.22   0.000     .8055218    .8820808
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,665
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8445646   .0173136    48.78   0.000     .8106216    .8785075
          2  |   .7743596    .012716    60.90   0.000     .7494303    .7992889
          3  |   .7629094   .0143188    53.28   0.000     .7348377    .7909811
          4  |   .8747072   .0164561    53.15   0.000     .8424455    .9069689
          5  |   .7863744   .0126016    62.40   0.000     .7616693    .8110796
          6  |   .7800449   .0147003    53.06   0.000     .7512253    .8088645
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0816552   .0224742     3.63   0.000     .0375952    .1257152
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         0.0817***
                     [0.0376,0.126]   
--------------------------------------
N                              4665   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,665
                                                F(7, 4664)        =       6.22
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0073
                                                Root MSE          =      .4053

                       (Std. err. adjusted for 4,665 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0301426   .0238819     1.26   0.207    -.0166772    .0769625
              |
          pk3 |
           2  |  -.0702049   .0214733    -3.27   0.001    -.1123028   -.0281071
           3  |  -.0816552   .0224742    -3.63   0.000    -.1257152   -.0375952
              |
treatment#pk3 |
         1 2  |  -.0181278   .0298517    -0.61   0.544    -.0766513    .0403957
         1 3  |  -.0130071   .0314794    -0.41   0.679    -.0747217    .0487074
              |
        round |
           2  |   -.011632    .015187    -0.77   0.444    -.0414057    .0181417
           3  |   .0118133   .0142919     0.83   0.409    -.0162055    .0398322
              |
        _cons |   .8438013   .0195256    43.22   0.000     .8055218    .8820808
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,665
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8445646   .0173136    48.78   0.000     .8106216    .8785075
          2  |   .7743596    .012716    60.90   0.000     .7494303    .7992889
          3  |   .7629094   .0143188    53.28   0.000     .7348377    .7909811
          4  |   .8747072   .0164561    53.15   0.000     .8424455    .9069689
          5  |   .7863744   .0126016    62.40   0.000     .7616693    .8110796
          6  |   .7800449   .0147003    53.06   0.000     .7512253    .8088645
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0946623   .0220783     4.29   0.000     .0513784    .1379463
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         0.0947***
                     [0.0514,0.138]   
--------------------------------------
N                              4665   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,665
                                                F(7, 4664)        =       6.22
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0073
                                                Root MSE          =      .4053

                       (Std. err. adjusted for 4,665 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0301426   .0238819     1.26   0.207    -.0166772    .0769625
              |
          pk3 |
           2  |  -.0702049   .0214733    -3.27   0.001    -.1123028   -.0281071
           3  |  -.0816552   .0224742    -3.63   0.000    -.1257152   -.0375952
              |
treatment#pk3 |
         1 2  |  -.0181278   .0298517    -0.61   0.544    -.0766513    .0403957
         1 3  |  -.0130071   .0314794    -0.41   0.679    -.0747217    .0487074
              |
        round |
           2  |   -.011632    .015187    -0.77   0.444    -.0414057    .0181417
           3  |   .0118133   .0142919     0.83   0.409    -.0162055    .0398322
              |
        _cons |   .8438013   .0195256    43.22   0.000     .8055218    .8820808
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,665
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .8445646   .0173136    48.78   0.000     .8106216    .8785075
          2  |   .7743596    .012716    60.90   0.000     .7494303    .7992889
          3  |   .7629094   .0143188    53.28   0.000     .7348377    .7909811
          4  |   .8747072   .0164561    53.15   0.000     .8424455    .9069689
          5  |   .7863744   .0126016    62.40   0.000     .7616693    .8110796
          6  |   .7800449   .0147003    53.06   0.000     .7512253    .8088645
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0130071   .0314794     0.41   0.679    -.0487074    .0747217
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1               0.0130   
                 (0.0315)   
----------------------------
N                    4665   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      3,882
                                                F(7, 3881)        =      12.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0217
                                                Root MSE          =     .49277

                       (Std. err. adjusted for 3,882 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0995578   .0425108    -2.34   0.019    -.1829033   -.0162122
              |
          pk3 |
           2  |   .0128651   .0283459     0.45   0.650    -.0427092    .0684394
           3  |   .1477699   .0289104     5.11   0.000     .0910889    .2044509
              |
treatment#pk3 |
         1 2  |   .1129502   .0491188     2.30   0.022      .016649    .2092514
         1 3  |   .0918868   .0495129     1.86   0.064    -.0051869    .1889605
              |
        round |
           2  |  -.0130888    .020001    -0.65   0.513    -.0523024    .0261247
           3  |  -.0204242   .0192182    -1.06   0.288     -.058103    .0172545
              |
        _cons |   .5007587   .0268231    18.67   0.000       .44817    .5533474
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 3,882
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4890234    .023924    20.44   0.000     .4421187    .5359282
          2  |   .5018886   .0152154    32.99   0.000     .4720577    .5317194
          3  |   .6367933   .0162175    39.27   0.000     .6049977     .668589
          4  |   .3894657   .0351542    11.08   0.000     .3205433    .4583881
          5  |    .515281    .019271    26.74   0.000     .4774987    .5530634
          6  |   .6291224   .0195329    32.21   0.000     .5908266    .6674181
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.1477699   .0289104    -5.11   0.000    -.2044509   -.0910889
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         -0.148***
                   [-0.204,-0.0911]   
--------------------------------------
N                              3882   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      3,882
                                                F(7, 3881)        =      12.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0217
                                                Root MSE          =     .49277

                       (Std. err. adjusted for 3,882 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0995578   .0425108    -2.34   0.019    -.1829033   -.0162122
              |
          pk3 |
           2  |   .0128651   .0283459     0.45   0.650    -.0427092    .0684394
           3  |   .1477699   .0289104     5.11   0.000     .0910889    .2044509
              |
treatment#pk3 |
         1 2  |   .1129502   .0491188     2.30   0.022      .016649    .2092514
         1 3  |   .0918868   .0495129     1.86   0.064    -.0051869    .1889605
              |
        round |
           2  |  -.0130888    .020001    -0.65   0.513    -.0523024    .0261247
           3  |  -.0204242   .0192182    -1.06   0.288     -.058103    .0172545
              |
        _cons |   .5007587   .0268231    18.67   0.000       .44817    .5533474
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 3,882
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4890234    .023924    20.44   0.000     .4421187    .5359282
          2  |   .5018886   .0152154    32.99   0.000     .4720577    .5317194
          3  |   .6367933   .0162175    39.27   0.000     .6049977     .668589
          4  |   .3894657   .0351542    11.08   0.000     .3205433    .4583881
          5  |    .515281    .019271    26.74   0.000     .4774987    .5530634
          6  |   .6291224   .0195329    32.21   0.000     .5908266    .6674181
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.2396567   .0402171    -5.96   0.000    -.3185054    -.160808
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         -0.240***
                    [-0.319,-0.161]   
--------------------------------------
N                              3882   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      3,882
                                                F(7, 3881)        =      12.56
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0217
                                                Root MSE          =     .49277

                       (Std. err. adjusted for 3,882 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0995578   .0425108    -2.34   0.019    -.1829033   -.0162122
              |
          pk3 |
           2  |   .0128651   .0283459     0.45   0.650    -.0427092    .0684394
           3  |   .1477699   .0289104     5.11   0.000     .0910889    .2044509
              |
treatment#pk3 |
         1 2  |   .1129502   .0491188     2.30   0.022      .016649    .2092514
         1 3  |   .0918868   .0495129     1.86   0.064    -.0051869    .1889605
              |
        round |
           2  |  -.0130888    .020001    -0.65   0.513    -.0523024    .0261247
           3  |  -.0204242   .0192182    -1.06   0.288     -.058103    .0172545
              |
        _cons |   .5007587   .0268231    18.67   0.000       .44817    .5533474
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 3,882
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4890234    .023924    20.44   0.000     .4421187    .5359282
          2  |   .5018886   .0152154    32.99   0.000     .4720577    .5317194
          3  |   .6367933   .0162175    39.27   0.000     .6049977     .668589
          4  |   .3894657   .0351542    11.08   0.000     .3205433    .4583881
          5  |    .515281    .019271    26.74   0.000     .4774987    .5530634
          6  |   .6291224   .0195329    32.21   0.000     .5908266    .6674181
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0918868   .0495129    -1.86   0.064    -.1889605    .0051869
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1              -0.0919   
                 (0.0495)   
----------------------------
N                    3882   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,336
                                                F(7, 4335)        =       2.42
                                                Prob > F          =     0.0180
                                                R-squared         =     0.0039
                                                Root MSE          =     .49531

                       (Std. err. adjusted for 4,336 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0097687   .0373939    -0.26   0.794    -.0830798    .0635425
              |
          pk3 |
           2  |   .0384201   .0282025     1.36   0.173    -.0168711    .0937114
           3  |   .0432244   .0290852     1.49   0.137    -.0137974    .1002463
              |
treatment#pk3 |
         1 2  |  -.0419693   .0436442    -0.96   0.336    -.1275342    .0435956
         1 3  |   .0266579    .044702     0.60   0.551    -.0609809    .1142968
              |
        round |
           2  |  -.0300888   .0190837    -1.58   0.115    -.0675025     .007325
           3  |   -.026325   .0181987    -1.45   0.148    -.0620038    .0093538
              |
        _cons |   .5586992   .0265066    21.08   0.000     .5067327    .6106657
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,336
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5393604   .0238596    22.61   0.000     .4925834    .5861374
          2  |   .5777805   .0150441    38.41   0.000     .5482865    .6072746
          3  |   .5825848   .0166173    35.06   0.000     .5500063    .6151633
          4  |   .5295917   .0288006    18.39   0.000     .4731277    .5860557
          5  |   .5260426   .0167219    31.46   0.000      .493259    .5588261
          6  |    .599474   .0180027    33.30   0.000     .5641795    .6347686
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0432244   .0290852    -1.49   0.137    -.1002463    .0137974
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0432   
                    [-0.100,0.0138]   
--------------------------------------
N                              4336   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,336
                                                F(7, 4335)        =       2.42
                                                Prob > F          =     0.0180
                                                R-squared         =     0.0039
                                                Root MSE          =     .49531

                       (Std. err. adjusted for 4,336 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0097687   .0373939    -0.26   0.794    -.0830798    .0635425
              |
          pk3 |
           2  |   .0384201   .0282025     1.36   0.173    -.0168711    .0937114
           3  |   .0432244   .0290852     1.49   0.137    -.0137974    .1002463
              |
treatment#pk3 |
         1 2  |  -.0419693   .0436442    -0.96   0.336    -.1275342    .0435956
         1 3  |   .0266579    .044702     0.60   0.551    -.0609809    .1142968
              |
        round |
           2  |  -.0300888   .0190837    -1.58   0.115    -.0675025     .007325
           3  |   -.026325   .0181987    -1.45   0.148    -.0620038    .0093538
              |
        _cons |   .5586992   .0265066    21.08   0.000     .5067327    .6106657
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,336
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5393604   .0238596    22.61   0.000     .4925834    .5861374
          2  |   .5777805   .0150441    38.41   0.000     .5482865    .6072746
          3  |   .5825848   .0166173    35.06   0.000     .5500063    .6151633
          4  |   .5295917   .0288006    18.39   0.000     .4731277    .5860557
          5  |   .5260426   .0167219    31.46   0.000      .493259    .5588261
          6  |    .599474   .0180027    33.30   0.000     .5641795    .6347686
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0698824   .0339707    -2.06   0.040    -.1364823   -.0032824
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0699*  
                  [-0.136,-0.00328]   
--------------------------------------
N                              4336   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,336
                                                F(7, 4335)        =       2.42
                                                Prob > F          =     0.0180
                                                R-squared         =     0.0039
                                                Root MSE          =     .49531

                       (Std. err. adjusted for 4,336 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |  -.0097687   .0373939    -0.26   0.794    -.0830798    .0635425
              |
          pk3 |
           2  |   .0384201   .0282025     1.36   0.173    -.0168711    .0937114
           3  |   .0432244   .0290852     1.49   0.137    -.0137974    .1002463
              |
treatment#pk3 |
         1 2  |  -.0419693   .0436442    -0.96   0.336    -.1275342    .0435956
         1 3  |   .0266579    .044702     0.60   0.551    -.0609809    .1142968
              |
        round |
           2  |  -.0300888   .0190837    -1.58   0.115    -.0675025     .007325
           3  |   -.026325   .0181987    -1.45   0.148    -.0620038    .0093538
              |
        _cons |   .5586992   .0265066    21.08   0.000     .5067327    .6106657
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,336
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5393604   .0238596    22.61   0.000     .4925834    .5861374
          2  |   .5777805   .0150441    38.41   0.000     .5482865    .6072746
          3  |   .5825848   .0166173    35.06   0.000     .5500063    .6151633
          4  |   .5295917   .0288006    18.39   0.000     .4731277    .5860557
          5  |   .5260426   .0167219    31.46   0.000      .493259    .5588261
          6  |    .599474   .0180027    33.30   0.000     .5641795    .6347686
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0266579    .044702    -0.60   0.551    -.1142968    .0609809
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1              -0.0267   
                 (0.0447)   
----------------------------
N                    4336   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,511
                                                F(7, 4510)        =       0.37
                                                Prob > F          =     0.9221
                                                R-squared         =     0.0006
                                                Root MSE          =      .4899

                       (Std. err. adjusted for 4,511 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0091932   .0351825     0.26   0.794    -.0597819    .0781682
              |
          pk3 |
           2  |   .0310383    .027904     1.11   0.266    -.0236671    .0857438
           3  |   .0155743   .0288631     0.54   0.590    -.0410116    .0721601
              |
treatment#pk3 |
         1 2  |  -.0075343   .0411885    -0.18   0.855     -.088284    .0732154
         1 3  |   .0066708   .0427924     0.16   0.876    -.0772232    .0905648
              |
        round |
           2  |  -.0084998   .0184863    -0.46   0.646    -.0447421    .0277425
           3  |  -.0034232   .0176958    -0.19   0.847    -.0381155    .0312692
              |
        _cons |   .5816673    .026176    22.22   0.000     .5303494    .6329851
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,511
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5777193   .0236249    24.45   0.000      .531403    .6240357
          2  |   .6087577   .0148634    40.96   0.000     .5796181    .6378972
          3  |   .5932936   .0165586    35.83   0.000     .5608306    .6257567
          4  |   .5869125   .0260777    22.51   0.000     .5357873    .6380376
          5  |   .6104166   .0154049    39.62   0.000     .5802155    .6406177
          6  |   .6091576   .0178752    34.08   0.000     .5741134    .6442018
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0155743   .0288631    -0.54   0.590    -.0721601    .0410116
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0156   
                   [-0.0722,0.0410]   
--------------------------------------
N                              4511   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,511
                                                F(7, 4510)        =       0.37
                                                Prob > F          =     0.9221
                                                R-squared         =     0.0006
                                                Root MSE          =      .4899

                       (Std. err. adjusted for 4,511 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0091932   .0351825     0.26   0.794    -.0597819    .0781682
              |
          pk3 |
           2  |   .0310383    .027904     1.11   0.266    -.0236671    .0857438
           3  |   .0155743   .0288631     0.54   0.590    -.0410116    .0721601
              |
treatment#pk3 |
         1 2  |  -.0075343   .0411885    -0.18   0.855     -.088284    .0732154
         1 3  |   .0066708   .0427924     0.16   0.876    -.0772232    .0905648
              |
        round |
           2  |  -.0084998   .0184863    -0.46   0.646    -.0447421    .0277425
           3  |  -.0034232   .0176958    -0.19   0.847    -.0381155    .0312692
              |
        _cons |   .5816673    .026176    22.22   0.000     .5303494    .6329851
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,511
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5777193   .0236249    24.45   0.000      .531403    .6240357
          2  |   .6087577   .0148634    40.96   0.000     .5796181    .6378972
          3  |   .5932936   .0165586    35.83   0.000     .5608306    .6257567
          4  |   .5869125   .0260777    22.51   0.000     .5357873    .6380376
          5  |   .6104166   .0154049    39.62   0.000     .5802155    .6406177
          6  |   .6091576   .0178752    34.08   0.000     .5741134    .6442018
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0222451   .0316106    -0.70   0.482    -.0842175    .0397272
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0222   
                   [-0.0842,0.0397]   
--------------------------------------
N                              4511   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,511
                                                F(7, 4510)        =       0.37
                                                Prob > F          =     0.9221
                                                R-squared         =     0.0006
                                                Root MSE          =      .4899

                       (Std. err. adjusted for 4,511 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0091932   .0351825     0.26   0.794    -.0597819    .0781682
              |
          pk3 |
           2  |   .0310383    .027904     1.11   0.266    -.0236671    .0857438
           3  |   .0155743   .0288631     0.54   0.590    -.0410116    .0721601
              |
treatment#pk3 |
         1 2  |  -.0075343   .0411885    -0.18   0.855     -.088284    .0732154
         1 3  |   .0066708   .0427924     0.16   0.876    -.0772232    .0905648
              |
        round |
           2  |  -.0084998   .0184863    -0.46   0.646    -.0447421    .0277425
           3  |  -.0034232   .0176958    -0.19   0.847    -.0381155    .0312692
              |
        _cons |   .5816673    .026176    22.22   0.000     .5303494    .6329851
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,511
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .5777193   .0236249    24.45   0.000      .531403    .6240357
          2  |   .6087577   .0148634    40.96   0.000     .5796181    .6378972
          3  |   .5932936   .0165586    35.83   0.000     .5608306    .6257567
          4  |   .5869125   .0260777    22.51   0.000     .5357873    .6380376
          5  |   .6104166   .0154049    39.62   0.000     .5802155    .6406177
          6  |   .6091576   .0178752    34.08   0.000     .5741134    .6442018
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0066708   .0427924    -0.16   0.876    -.0905648    .0772232
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1             -0.00667   
                 (0.0428)   
----------------------------
N                    4511   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,476
                                                F(7, 4475)        =       3.18
                                                Prob > F          =     0.0023
                                                R-squared         =     0.0052
                                                Root MSE          =     .41054

                       (Std. err. adjusted for 4,476 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .050025   .0300905     1.66   0.096    -.0089673    .1090172
              |
          pk3 |
           2  |   .0318589   .0245557     1.30   0.195    -.0162825    .0800003
           3  |   .0741565   .0247818     2.99   0.003     .0255719    .1227411
              |
treatment#pk3 |
         1 2  |  -.0422045   .0352884    -1.20   0.232    -.1113873    .0269783
         1 3  |   -.059176     .03592    -1.65   0.100     -.129597     .011245
              |
        round |
           2  |   .0344229   .0158376     2.17   0.030     .0033733    .0654724
           3  |   .0492261   .0149871     3.28   0.001     .0198439    .0786082
              |
        _cons |   .7094711   .0232139    30.56   0.000     .6639604    .7549818
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,476
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7386351   .0209952    35.18   0.000     .6974742    .7797959
          2  |   .7704939   .0127609    60.38   0.000     .7454762    .7955117
          3  |   .8127916   .0131665    61.73   0.000     .7869787    .8386044
          4  |     .78866   .0215784    36.55   0.000     .7463558    .8309643
          5  |   .7783144   .0133269    58.40   0.000     .7521871    .8044418
          6  |   .8036405   .0145637    55.18   0.000     .7750885    .8321925
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0741565   .0247818    -2.99   0.003    -.1227411   -.0255719
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0742** 
                   [-0.123,-0.0256]   
--------------------------------------
N                              4476   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,476
                                                F(7, 4475)        =       3.18
                                                Prob > F          =     0.0023
                                                R-squared         =     0.0052
                                                Root MSE          =     .41054

                       (Std. err. adjusted for 4,476 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .050025   .0300905     1.66   0.096    -.0089673    .1090172
              |
          pk3 |
           2  |   .0318589   .0245557     1.30   0.195    -.0162825    .0800003
           3  |   .0741565   .0247818     2.99   0.003     .0255719    .1227411
              |
treatment#pk3 |
         1 2  |  -.0422045   .0352884    -1.20   0.232    -.1113873    .0269783
         1 3  |   -.059176     .03592    -1.65   0.100     -.129597     .011245
              |
        round |
           2  |   .0344229   .0158376     2.17   0.030     .0033733    .0654724
           3  |   .0492261   .0149871     3.28   0.001     .0198439    .0786082
              |
        _cons |   .7094711   .0232139    30.56   0.000     .6639604    .7549818
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,476
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7386351   .0209952    35.18   0.000     .6974742    .7797959
          2  |   .7704939   .0127609    60.38   0.000     .7454762    .7955117
          3  |   .8127916   .0131665    61.73   0.000     .7869787    .8386044
          4  |     .78866   .0215784    36.55   0.000     .7463558    .8309643
          5  |   .7783144   .0133269    58.40   0.000     .7521871    .8044418
          6  |   .8036405   .0145637    55.18   0.000     .7750885    .8321925
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0149805   .0260367    -0.58   0.565    -.0660254    .0360644
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0150   
                   [-0.0660,0.0361]   
--------------------------------------
N                              4476   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,476
                                                F(7, 4475)        =       3.18
                                                Prob > F          =     0.0023
                                                R-squared         =     0.0052
                                                Root MSE          =     .41054

                       (Std. err. adjusted for 4,476 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .050025   .0300905     1.66   0.096    -.0089673    .1090172
              |
          pk3 |
           2  |   .0318589   .0245557     1.30   0.195    -.0162825    .0800003
           3  |   .0741565   .0247818     2.99   0.003     .0255719    .1227411
              |
treatment#pk3 |
         1 2  |  -.0422045   .0352884    -1.20   0.232    -.1113873    .0269783
         1 3  |   -.059176     .03592    -1.65   0.100     -.129597     .011245
              |
        round |
           2  |   .0344229   .0158376     2.17   0.030     .0033733    .0654724
           3  |   .0492261   .0149871     3.28   0.001     .0198439    .0786082
              |
        _cons |   .7094711   .0232139    30.56   0.000     .6639604    .7549818
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,476
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7386351   .0209952    35.18   0.000     .6974742    .7797959
          2  |   .7704939   .0127609    60.38   0.000     .7454762    .7955117
          3  |   .8127916   .0131665    61.73   0.000     .7869787    .8386044
          4  |     .78866   .0215784    36.55   0.000     .7463558    .8309643
          5  |   .7783144   .0133269    58.40   0.000     .7521871    .8044418
          6  |   .8036405   .0145637    55.18   0.000     .7750885    .8321925
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |    .059176     .03592     1.65   0.100     -.011245     .129597
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1               0.0592   
                 (0.0359)   
----------------------------
N                    4476   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,651
                                                F(7, 4650)        =       1.35
                                                Prob > F          =     0.2234
                                                R-squared         =     0.0019
                                                Root MSE          =     .42083

                       (Std. err. adjusted for 4,651 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0719811   .0286454     2.51   0.012     .0158224    .1281397
              |
          pk3 |
           2  |   .0107241   .0246535     0.43   0.664    -.0376085    .0590567
           3  |   .0376994   .0251005     1.50   0.133    -.0115095    .0869084
              |
treatment#pk3 |
         1 2  |  -.0571891   .0340906    -1.68   0.093     -.124023    .0096447
         1 3  |  -.0826471   .0350841    -2.36   0.019    -.1514287   -.0138656
              |
        round |
           2  |   .0035804   .0155497     0.23   0.818    -.0269045    .0340652
           3  |  -.0046158   .0149753    -0.31   0.758    -.0339746    .0247429
              |
        _cons |   .7443932   .0229665    32.41   0.000     .6993681    .7894183
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,651
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7437796   .0208863    35.61   0.000     .7028325    .7847267
          2  |   .7545037   .0130991    57.60   0.000     .7288232    .7801843
          3  |   .7814791   .0139203    56.14   0.000     .7541886    .8087695
          4  |   .8157607    .019614    41.59   0.000      .777308    .8542134
          5  |   .7692956   .0130719    58.85   0.000     .7436686    .7949227
          6  |    .770813   .0147233    52.35   0.000     .7419483    .7996777
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0376994   .0251005    -1.50   0.133    -.0869084    .0115095
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0377   
                   [-0.0869,0.0115]   
--------------------------------------
N                              4651   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,651
                                                F(7, 4650)        =       1.35
                                                Prob > F          =     0.2234
                                                R-squared         =     0.0019
                                                Root MSE          =     .42083

                       (Std. err. adjusted for 4,651 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0719811   .0286454     2.51   0.012     .0158224    .1281397
              |
          pk3 |
           2  |   .0107241   .0246535     0.43   0.664    -.0376085    .0590567
           3  |   .0376994   .0251005     1.50   0.133    -.0115095    .0869084
              |
treatment#pk3 |
         1 2  |  -.0571891   .0340906    -1.68   0.093     -.124023    .0096447
         1 3  |  -.0826471   .0350841    -2.36   0.019    -.1514287   -.0138656
              |
        round |
           2  |   .0035804   .0155497     0.23   0.818    -.0269045    .0340652
           3  |  -.0046158   .0149753    -0.31   0.758    -.0339746    .0247429
              |
        _cons |   .7443932   .0229665    32.41   0.000     .6993681    .7894183
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,651
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7437796   .0208863    35.61   0.000     .7028325    .7847267
          2  |   .7545037   .0130991    57.60   0.000     .7288232    .7801843
          3  |   .7814791   .0139203    56.14   0.000     .7541886    .8087695
          4  |   .8157607    .019614    41.59   0.000      .777308    .8542134
          5  |   .7692956   .0130719    58.85   0.000     .7436686    .7949227
          6  |    .770813   .0147233    52.35   0.000     .7419483    .7996777
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0449477   .0245291     1.83   0.067    -.0031409    .0930363
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         0.0449   
                  [-0.00314,0.0930]   
--------------------------------------
N                              4651   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,651
                                                F(7, 4650)        =       1.35
                                                Prob > F          =     0.2234
                                                R-squared         =     0.0019
                                                Root MSE          =     .42083

                       (Std. err. adjusted for 4,651 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |   .0719811   .0286454     2.51   0.012     .0158224    .1281397
              |
          pk3 |
           2  |   .0107241   .0246535     0.43   0.664    -.0376085    .0590567
           3  |   .0376994   .0251005     1.50   0.133    -.0115095    .0869084
              |
treatment#pk3 |
         1 2  |  -.0571891   .0340906    -1.68   0.093     -.124023    .0096447
         1 3  |  -.0826471   .0350841    -2.36   0.019    -.1514287   -.0138656
              |
        round |
           2  |   .0035804   .0155497     0.23   0.818    -.0269045    .0340652
           3  |  -.0046158   .0149753    -0.31   0.758    -.0339746    .0247429
              |
        _cons |   .7443932   .0229665    32.41   0.000     .6993681    .7894183
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,651
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .7437796   .0208863    35.61   0.000     .7028325    .7847267
          2  |   .7545037   .0130991    57.60   0.000     .7288232    .7801843
          3  |   .7814791   .0139203    56.14   0.000     .7541886    .8087695
          4  |   .8157607    .019614    41.59   0.000      .777308    .8542134
          5  |   .7692956   .0130719    58.85   0.000     .7436686    .7949227
          6  |    .770813   .0147233    52.35   0.000     .7419483    .7996777
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0826471   .0350841     2.36   0.019     .0138656    .1514287
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1               0.0826*  
                 (0.0351)   
----------------------------
N                    4651   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,544
                                                F(7, 4543)        =       6.34
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0096
                                                Root MSE          =      .4979

                       (Std. err. adjusted for 4,544 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .043022   .0352313     1.22   0.222    -.0260484    .1120924
              |
          pk3 |
           2  |   .0591469   .0280748     2.11   0.035     .0041067    .1141871
           3  |   .1292174   .0289272     4.47   0.000     .0725061    .1859288
              |
treatment#pk3 |
         1 2  |   .0004325   .0415259     0.01   0.992    -.0809783    .0818434
         1 3  |  -.0363439   .0428399    -0.85   0.396     -.120331    .0476432
              |
        round |
           2  |  -.0466701   .0187185    -2.49   0.013    -.0833676   -.0099727
           3  |  -.0427228   .0178964    -2.39   0.017    -.0778085   -.0076371
              |
        _cons |   .4539623   .0262133    17.32   0.000     .4025714    .5053532
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,544
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4233509   .0235971    17.94   0.000      .377089    .4696128
          2  |   .4824977   .0152245    31.69   0.000     .4526503    .5123452
          3  |   .5525683   .0167066    33.07   0.000     .5198152    .5853214
          4  |   .4663729   .0261734    17.82   0.000     .4150603    .5176854
          5  |   .5259523   .0158352    33.21   0.000     .4949075    .5569971
          6  |   .5592464   .0177537    31.50   0.000     .5244404    .5940523
------------------------------------------------------------------------------

        lc_1:  1._at-3._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.1292174   .0289272    -4.47   0.000    -.1859288   -.0725061
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                         -0.129***
                   [-0.186,-0.0725]   
--------------------------------------
N                              4544   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,544
                                                F(7, 4543)        =       6.34
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0096
                                                Root MSE          =      .4979

                       (Std. err. adjusted for 4,544 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .043022   .0352313     1.22   0.222    -.0260484    .1120924
              |
          pk3 |
           2  |   .0591469   .0280748     2.11   0.035     .0041067    .1141871
           3  |   .1292174   .0289272     4.47   0.000     .0725061    .1859288
              |
treatment#pk3 |
         1 2  |   .0004325   .0415259     0.01   0.992    -.0809783    .0818434
         1 3  |  -.0363439   .0428399    -0.85   0.396     -.120331    .0476432
              |
        round |
           2  |  -.0466701   .0187185    -2.49   0.013    -.0833676   -.0099727
           3  |  -.0427228   .0178964    -2.39   0.017    -.0778085   -.0076371
              |
        _cons |   .4539623   .0262133    17.32   0.000     .4025714    .5053532
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,544
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4233509   .0235971    17.94   0.000      .377089    .4696128
          2  |   .4824977   .0152245    31.69   0.000     .4526503    .5123452
          3  |   .5525683   .0167066    33.07   0.000     .5198152    .5853214
          4  |   .4663729   .0261734    17.82   0.000     .4150603    .5176854
          5  |   .5259523   .0158352    33.21   0.000     .4949075    .5569971
          6  |   .5592464   .0177537    31.50   0.000     .5244404    .5940523
------------------------------------------------------------------------------

        lc_1:  4._at-6._at = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |  -.0928735   .0316291    -2.94   0.003     -.154882    -.030865
------------------------------------------------------------------------------

--------------------------------------
                                (1)   
                                  .   
--------------------------------------
lc_1                        -0.0929** 
                   [-0.155,-0.0309]   
--------------------------------------
N                              4544   
--------------------------------------
95% confidence intervals in brackets
* p<0.05, ** p<0.01, *** p<0.001

Linear regression                               Number of obs     =      4,544
                                                F(7, 4543)        =       6.34
                                                Prob > F          =     0.0000
                                                R-squared         =     0.0096
                                                Root MSE          =      .4979

                       (Std. err. adjusted for 4,544 clusters in respondent_id)
-------------------------------------------------------------------------------
              |               Robust
      support | Coefficient  std. err.      t    P>|t|     [95% conf. interval]
--------------+----------------------------------------------------------------
  1.treatment |    .043022   .0352313     1.22   0.222    -.0260484    .1120924
              |
          pk3 |
           2  |   .0591469   .0280748     2.11   0.035     .0041067    .1141871
           3  |   .1292174   .0289272     4.47   0.000     .0725061    .1859288
              |
treatment#pk3 |
         1 2  |   .0004325   .0415259     0.01   0.992    -.0809783    .0818434
         1 3  |  -.0363439   .0428399    -0.85   0.396     -.120331    .0476432
              |
        round |
           2  |  -.0466701   .0187185    -2.49   0.013    -.0833676   -.0099727
           3  |  -.0427228   .0178964    -2.39   0.017    -.0778085   -.0076371
              |
        _cons |   .4539623   .0262133    17.32   0.000     .4025714    .5053532
-------------------------------------------------------------------------------

Predictive margins                                       Number of obs = 4,544
Model VCE: Robust

Expression: Linear prediction, predict()
1._at: treatment = 0
       pk3       = 1
2._at: treatment = 0
       pk3       = 2
3._at: treatment = 0
       pk3       = 3
4._at: treatment = 1
       pk3       = 1
5._at: treatment = 1
       pk3       = 2
6._at: treatment = 1
       pk3       = 3

------------------------------------------------------------------------------
             |            Delta-method
             |     Margin   std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
         _at |
          1  |   .4233509   .0235971    17.94   0.000      .377089    .4696128
          2  |   .4824977   .0152245    31.69   0.000     .4526503    .5123452
          3  |   .5525683   .0167066    33.07   0.000     .5198152    .5853214
          4  |   .4663729   .0261734    17.82   0.000     .4150603    .5176854
          5  |   .5259523   .0158352    33.21   0.000     .4949075    .5569971
          6  |   .5592464   .0177537    31.50   0.000     .5244404    .5940523
------------------------------------------------------------------------------

        lc_1:  (4._at-6._at)-(1._at-3._at) = 0

------------------------------------------------------------------------------
             | Coefficient  Std. err.      t    P>|t|     [95% conf. interval]
-------------+----------------------------------------------------------------
        lc_1 |   .0363439   .0428399     0.85   0.396    -.0476432     .120331
------------------------------------------------------------------------------

----------------------------
                      (1)   
                        .   
----------------------------
lc_1               0.0363   
                 (0.0428)   
----------------------------
N                    4544   
----------------------------
Standard errors in parentheses
* p<0.05, ** p<0.01, *** p<0.001

. mat ci = ci1\ci2\ci3\ci4\ci5\ci6\ci7\ci8\cit1\cit2\cit3\cit4\cit5\cit6\cit7\cit8

. 
. mat list ci

ci[16,4]
          active:     active:     active:     active:
               b        ci_l        ci_u           p
lc_1  -.02306431  -.06479193    .0186633   .27858657
lc_1   .08165519   .03759518    .1257152   .00028284
lc_1  -.14776989  -.20445093  -.09108885   3.354e-07
lc_1  -.04322442  -.10024629   .01379745   .13731716
lc_1  -.01557429  -.07216013   .04101156   .58950551
lc_1  -.07415652   -.1227411  -.02557193   .00278325
lc_1  -.03769945  -.08690835   .01150946   .13318066
lc_1  -.12921744  -.18592882  -.07250605   8.126e-06
lc_1   .03952565  -.00192849   .08097979   .06164779
lc_1   .09466233   .05137838   .13794629   .00001843
lc_1  -.23965669   -.3185054  -.16080799   2.762e-09
lc_1  -.06988235  -.13648232  -.00328239   .03973288
lc_1  -.02224511  -.08421747   .03972724   .48164213
lc_1  -.01498048  -.06602537   .03606441   .56507731
lc_1   .04494768   -.0031409   .09303627    .0669519
lc_1   -.0928735  -.15488199  -.03086501    .0033379

. 
. sort respondent_id topic

. 
. keep in 1/16
(38,464 observations deleted)

. 
. keep topic topic2

. 
. g treatment=0 

. replace treatment=1 in 9/16
(8 real changes made)

. 
. svmat double ci, name(temp)

. ren temp1 margin

. ren temp2 lb

. ren temp3 ub

. drop temp4

. 
. mat did = did1\did2\did3\did4\did5\did6\did7\did8

. 
. mat list did

did[8,3]
          active:     active:     active:
               b          se           p
lc_1   .06258997   .02998834   .03693087
lc_1   .01300714   .03147943   .67948284
lc_1   -.0918868   .04951288   .06355551
lc_1  -.02665793   .04470203   .55097443
lc_1  -.00667083   .04279236   .87612808
lc_1   .05917603   .03592002   .09953811
lc_1   .08264713   .03508412   .01852991
lc_1   .03634393   .04283992   .39627892

. 
. svmat double did, name(temp)

. ren temp1 did_est

. ren temp2 did_se

. ren temp3 did_p

. 
. label define topiclab_alt2 1 "Infrastructure" 2 "Min wage" 3 "Estate tax" 4 "CG tax" 5 "Bord
> er sec" 6 "Transgender" 7 "Abortion" 8 "Covid"

. label values topic topiclab_alt2

. 
. sort topic treatment

. g hline=ub+0.03 if treatment==1
(8 missing values generated)

. replace hline=hline[_n+1] if hline==.
(8 real changes made)

. 
. g temp=ub+0.03 if treatment==0
(8 missing values generated)

. replace temp=temp[_n-1] if temp==.
(8 real changes made)

. 
. replace hline=temp if topic==3 | topic==4 | topic==5
(6 real changes made)

. 
. drop temp 

. 
. g dropl=hline

. replace dropl=hline+0.0012 if treatment==0
(8 real changes made)

. replace dropl=hline-0.01 if treatment==1
(8 real changes made)

. 
. ** labels
. * slope coefficient
. su did_est if treatment==0 & topic==1

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1      .06259           .     .06259     .06259

. local margin_3f : di %9.3f r(mean)

. di `margin_3f'
.063

. g dydx_t="`margin_3f'" if treatment==0 & topic==1
(15 missing values generated)

. 
. foreach n of numlist 2/8 {
  2.         su did_est if treatment==0 & topic== `n'
  3.         local margin_3f : di %9.3f r(mean)
  4.         replace dydx_t="`margin_3f'" if treatment==0 & topic== `n'
  5. }

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1    .0130071           .   .0130071   .0130071
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1   -.0918868           .  -.0918868  -.0918868
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1   -.0266579           .  -.0266579  -.0266579
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1   -.0066708           .  -.0066708  -.0066708
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1     .059176           .    .059176    .059176
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1    .0826471           .   .0826471   .0826471
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
     did_est |          1    .0363439           .   .0363439   .0363439
(1 real change made)

. 
. 
. * se
. su did_se if treatment==0 & topic== 1

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0299883           .   .0299883   .0299883

. local margin_3f : di %9.3f r(mean)

. di `margin_3f'
.03

. g se_t="(`margin_3f')" if treatment==0 & topic==1
(15 missing values generated)

. 
. 
. foreach n of numlist 2/8 {
  2.         su did_se if treatment==0 & topic== `n'
  3.         local margin_3f : di %9.3f r(mean)
  4.         replace se_t="(`margin_3f')" if treatment==0 & topic== `n'
  5. }

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0314794           .   .0314794   .0314794
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0495129           .   .0495129   .0495129
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1     .044702           .    .044702    .044702
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0427924           .   .0427924   .0427924
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1      .03592           .     .03592     .03592
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0350841           .   .0350841   .0350841
(1 real change made)

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      did_se |          1    .0428399           .   .0428399   .0428399
(1 real change made)

. 
. replace se_t = subinstr(se_t, " ", "", .)
(8 real changes made)

. 
. g lab_t=dydx_t+" "+se_t

. 
. g z=0

. g o=1

. g x=0.08

. g x2=0.14

. g y_h=hline+0.04

. g y_h2=hline+0.02

. 
. twoway ///
> (scatter margin treatment if treatment==0, msymbol(O) mcolor(black) lcolor(black)) ///
> (scatter margin treatment if treatment==1, msymbol(O) mcolor(gray) lcolor(gray) lpattern(sol
> id)) ///
> (rspike lb ub treatment if treatment==0,  lcolor(black)  lwidth(medthick)) ///
> (rspike lb ub treatment if treatment==1, lcolor(gray)  lwidth(medthick)) ///
>  (line  hline treatment ,  lw(medthick)  mc(none) lc(black) lp(solid)) ///
>  (line  dropl z,  lw(medthick)  mc(none) lc(black) lp(solid)) ///
>  (line  dropl o,  lw(medthick)  mc(none) lc(black) lp(solid)) ///
>  (scatter y_h x2, ms(none) mla(dydx_t) mlabcolor(black) mlabsize(medsmall)) ///
>  (scatter y_h2 x, ms(none) mla(se_t) mlabcolor(black) mlabsize(medsmall)), ///
> by(topic, r(1) legend(off) note("") title("", size(large) position(11))) subtitle(,  size(me
> dsmall)) ///
> scheme(plotplain) yline(.5) xtick(0 1) xscale(r(-0.2 1.2)) xlabel(0 "Control" 1 "Treat.", la
> bsize(medsmall)  labcolor(gray) tlength(0.5)) ylabel(-0.3(.1)0.3, labsize(medsmall) labcolor
> (gray) format(%3.1f)) xtitle("", size(medsmall) color(gray)) ytitle("", size(medsmall) color
> (gray)) xsize(8)

. 
. graph export "figure6.pdf",  replace
file /Users/qsg999/Desktop/PSRM_replication_material/temp/figure6.pdf saved as PDF format

. 
. restore

. 
. log close 
      name:  <unnamed>
       log:  /Users/qsg999/Desktop/PSRM_replication_material/Main_analysis.log
  log type:  text
 closed on:  18 Jun 2024, 09:03:05
----------------------------------------------------------------------------------------------
